Artificial Intelligence Creates Functional Bacteriophage Genomes, Countering Antibiotic Resistance Crisis
Stanford University's Assistant Professor of Chemical Engineering, Brian Hie, PhD, and his team have made a groundbreaking discovery in the field of synthetic biology. In a collaborative study with Nvidia, they have demonstrated the first end-to-end generative design of 16 complete, functional, and evolutionarily novel bacteriophage genomes.
The research, led by Samuel King, a PhD candidate in the Hie lab and first author of the preprint, was conducted using the historic bacteriophage ΦX174, first sequenced by Nobel Laureate Frederick Sanger, and first chemically synthesized by Craig Venter, as a design template. The team employed the Evo series of foundation models, with Evo 2, described as the largest publicly available AI model for biology to date, at the forefront.
The DNA sequences of living organisms contain instructions for life at multiple scales, including individual pathways and whole genomes. The tightly orchestrated interactions between multiple genes, regulatory elements, and recognition sequences in genomes are complex and a challenge to recreate. However, the AI-designed genomes have shown promise in informing a phage-based strategy to target multi-drug resistant bacteria that impact significant percentages of agricultural crops worldwide.
The study, published without specifying when and by whom the first end-to-end generatively designed, functional, and evolutionarily novel bacteriophage genomes were initially published, used the historic ΦX174 as a starting point. Out of approximately 300 designs that were experimentally tested, 16 phages emerged as viable with substantial evolutionary diversity. Some of the generated phages showed increased fitness relative to ΦX174 in growth competitions and lysis kinetics.
Hie's motivation is to build genome foundation models for generative design to engineer more complex biological functions. This research is a significant step towards this goal, marking the first experimental validation of the whole genome design concept.
The Virtual Cell Challenge, a public competition for machine learning models predicting how cells will respond to genetic perturbations, was recently launched by Arc Institute. This competition, along with future directions of the study aiming to design larger phage genomes and other genomic systems, such as operons in bacteria, will continue to push the boundaries of what is possible in the realm of synthetic biology.
Read also:
- Abu Dhabi initiative for comprehensive genetic screening, aiming to diagnose over 800 conditions and enhance the health of future generations in the UAE.
- Elderly shingles: Recognizing symptoms, potential problems, and available treatments
- Exploring the Reasons, Purposes, and Enigmas of Hiccups: Delving into Their Origins, Roles, and Unsolved Aspects
- Various forms of cataracts include nuclear, pediatric, traumatic, and additional types
